cho.sh
2,238 notes
261021261021Toward Autopoiesis26F6DF오토포이에시스를 향해84A5EE차세대 디지털 브레인 만들기C24273261019261019261018261018디지털 시대의 언어 다양성을 위한 선언문507C2CManifesto for Linguistic Diversity in the Digital Age6B5B52261012261012코리지빌리티1CAB16261005261005261004261004Mesh EngineeringF27D1B그물망 엔지니어링881E3BThe devil is in the details94F308260923260923260919260919260917260917In the Agentic Era, should we become a dumb pipe?D815AC260916260916Coscientist6975BA260915260915Coscientific Swarm5F70E8협력적 연구 군체B9E26A
Service Shimming

Service Shimming

Warning

Over a year old. May be outdated

In computer programming, shimming means transparently intercepting, modifying, and redirecting API calls. A smooth shim is hardly noticeable. Think Mission Impossible. You order food at a restaurant, but Ethan Hunt intercepts your call, cooks the food, and delivers it himself.

People really don't care about how developers implement services. If you withdraw some money from an ATM, you don't care if it's an actual ATM, a secret terminator trying to assassinate you, or actually, there's a person inside a machine counting the money for you. The important thing is giving the illusion that everything is working. If this is true, engineering might not be as crucial as we developers think. Maybe is more critical. Here are a few examples:

  • Edison faked his invention of the light bulb for more than 4 years.
  • The first iPhone failed to integrate individual apps, so Jobs had to prepare separate iPhones for calls, music, maps, etc. and switch between them, pretending it was the same device.
  • Tesla is also a similar example. Elon always makes bold claims but never delivers on that date. Where is our Cybertruck?
  • Naver (≈ Korean Google), for a while, failed to make a News automation service, so the CEO visited press organizations every morning to ask permission to distribute the articles.
  • Woowa Bros (≈ Korean Doordash) failed to make an order relay system, so they had to call the restaurant one by one whenever they got an order request.
  • (≈ Korean Paypal) failed to make the transaction system, so the CEO visited the bank three times a day to bulk-process pending Venmo requests.

Theranos, from the "Bad blood" incident, followed the same flow. They claimed to diagnose NN amount of tests with a drop of blood, but the device never worked, and they operated a full-scale diagnosis lab to fake as if the results were coming from their flagship machine. Soon they were exposed and ended in a fiasco. How would history change if they had succeeded in the end? One more thing. 4 years in the 20th century and 21st century is extremely different in velocity.

How far can we accept fake it until you make it?

Backlinks5